swarm-architecture-design

Adversarial architecture decision-making uses parallel advocate agents to build the strongest cases for competing technology options, synthesizing evidence and constraints to improve bias resistance and decision quality.

Swarm analysis in architecture research triages decision areas into swarm-worthy and sequential categories, automatically spawning advocate agents per viable option to evaluate technologies like databases and messaging systems in parallel.

If the Agent tool is unavailable, the architecture design skill gracefully falls back to the base architecture skill and logs the fallback event in the research journal to ensure continuity.

Evaluating database options for a multi-tenant system involves spawning parallel swarm advocates per option to gather evidence, scan cross-decision conflicts, and synthesize consolidated recommendations based on strong isolation requirements.

Adversarial swarm analysis is used when you have two or more viable technology options for an architecture decision and need evidence-based synthesis to compare trade-offs and improve bias resistance.

A limitation of using advocate agents is the dependency on the Agent tool; without it, the system cannot spawn parallel advocates and must fall back to the base architecture skill, reducing adversarial capability.